A major goal of HIV vaccine candidates is to elicit broadly neutralizing antibodies. Many features of HIV envelope glycoproteins (Env), the target of neutralizing antibodies, make this a difficult goal. These features include poor exposure of conserved regions of Env, better exposure of variable regions and a dense carbohydrate covering. Nevertheless, a small number of broadly neutralizing monoclonal antibodies (mAbs) directed against conserved epitopes have been isolated and they represent potential tools for vaccine design. This application concentrates on one of these, mAb b12, and, in particular, on attempts to design molecules that focus antibody responses to the 1012 epitope, which overlaps the conserved CD4 binding site on gp120. We have solved the structure of b12 and characterized its interaction with Env gp120 in some detail by modeling, mutagenesis and other studies. We have shown how gp120 can be engineered, notably by the introduction of extra carbohydrate groups, to preferentially bind b12 relative to a panel of non- and weakly neutralizing mAbs. Here we propose to investigate the immunogenicity of engineered gp120s in animal models and to use the resulting data to improve immunogen design. We also propose to investigate the ability of the engineering strategy to immunofocus responses to the b12 epitope on recombinant trimeric Env and virus-like particles (VLPs). Finally, in a conceptually different approach, we will chemically modify gp120 bound to b12 and then attach immuno-inert large moieties, including carbohydrates, to mask the regions not contacted by b12. Immunization with this molecule should, in principle, be akin to immunizing with the footprint of b12. [unreadable] [unreadable] [unreadable]